Biomimetic model of the outer plexiform layer by incorporating memristive devices

Abstract

In this paper we present a biorealistic model for the first part of the early vision processing by incorporating memristive nanodevices. The architecture of the proposed network is based on the organisation and functioning of the outer plexiform layer (OPL) in the vertebrate retina. We demonstrate that memristive devices are indeed a valuable building block for neuromorphic architectures, as their highly non-linear and adaptive response could be exploited for establishing ultra-dense networks with similar dynamics to their biological counterparts. We particularly show that hexagonal memristive grids can be employed for faithfully emulating the smoothing-effect occurring at the OPL for enhancing the dynamic range of the system. In addition, we employ a memristor-based thresholding scheme for detecting the edges of grayscale images, while the proposed system is also evaluated for its adaptation and fault tolerance capacity against different light or noise conditions as well as distinct device yields.Over the past years, the performance and efficiency of biological systems has inspired many researchers and engineers, giving birth to the emerging fields of biomimetics [1] and bioinspiration [2]. The human retina is anything but a simple passive relay station, as it preprocesses and compresses all sensed information through an immensely complex neuronal network that on average contains 4.6 million cones, 92 million rods [3] and 1 million ganglion [4]. If we take into account the remaining bipolar, horizontal and amacrine cells and the fact that they are highly interconnected we get a parallel network of grand complexity. This complexity is further elucidated in recent studies where evidence is provided that at least ten parallel signals arise from a single visual point [5] and we get a much more accurate view from the functioning of the vertebrate retina. In fact it is nowadays believed that although biological systems are

Memristive Biosensors for PSA-IgM Detection

“…Memristive devices present a rapidly increasing interest in recent years thanks to the different application they may be used for. Indicatively, memristive systems enable new possibilities for computation and non-volatile memory storage as, for example, the construction of memristor-based digital logic circuits [13], applications regarding artificial synapses [14][15][16], resistive RAM (ReRAM) memories [17], and generic memristive structure (GMS) for 3D FPGA applications have been proposed [18]. Memristive silicon nanowire devices functionalized with antibody films are recently introduced as memristive biosensors for biosensing purposes [19,20].…”

“Bio-functionalization study of Memristive-Biosensors for early detection of prostate cancer”

“…The hysteresis appearing in the current-to-voltage characteristics has been attributed to a wide range of possible phenomena, including molecular redox events, and metal filament formation and destruction. Memristive devices have already been used in many applications, since memristors enable new possibilities for computation and non-volatile memory storage as for example the construction of memristor-based digital logic circuits [14], applications regarding artificial synapses [15][16][17], Resistive RAM (ReRAM) memories [18] and Generic Memristive Structure (GMS) for 3-D FPGA applications have been proposed [19]. Moreover, memristive silicon nanowire devices functionalized with antibody films are recently introduced as Memristive-Biosensors for bio sensing purposes taking advantage of their memristive electrical characteristics [20,21].…”

Resistive switching characteristics of indium‐tin‐oxide thin film devices

“…In 2008, a group from HP initiated a new perspective in ReRAM by fabricating nanoscale TiO 2 two‐terminal device and correlating its dynamics to the memristor described mathematically by Chua a phenomenon which has however been unconsciously observed for more than two centuries . Since HP's original work, research interest is increasing towards developing reliable devices that can be exploited in practical applications such as artificial synapses, non‐volatile memory, neuromorphic computing and image processing . Nonetheless, the functional properties of distinct systems are still under debate because of the ambiguity of the conducting mechanism supported within the employed active core, namely, chemical reaction based mechanisms involves mobile species or physical changes such as magnetic tunnel junctions and phase change of the devices' functional core .…”